Dataset Viewer
Auto-converted to Parquet Duplicate
Search is not available for this dataset
row
int64
19
1,000k
loss
listlengths
54
54
19
[ 0.48996055126190186, 0.08315947651863098, 1.185167908668518, 0.7175381779670715, 0.7492619752883911, 0.39704081416130066, 1.1239914894104004, 0.45560967922210693, 0.2546129822731018, 0.16126039624214172, 0.15904997289180756, 1.106515645980835, 0.6376010179519653, 0.47192084789276123, 0.2...
150
[ 2.2576651573181152, 1.832277774810791, 2.254758834838867, 1.0651841163635254, 2.0158133506774902, 2.86967134475708, 1.7468936443328857, 1.5051778554916382, 1.4827688932418823, 1.3864386081695557, 1.4015898704528809, 1.1768923997879028, 1.9890638589859009, 1.7520840167999268, 2.7701027393...
188
[ 0.04157457873225212, 0.045340731739997864, 0.04706236347556114, 0.03986649215221405, 0.08272355794906616, 0.046618152409791946, 0.022755103185772896, 0.028834575787186623, 0.03736938536167145, 0.02120513841509819, 0.03470369055867195, 0.025850284844636917, 0.06885264068841934, 0.0330735482...
219
[ 1.7058615684509277, 0.9044452905654907, 1.4710230827331543, 0.9830461740493774, 1.8850185871124268, 1.2415738105773926, 0.8816855549812317, 1.934943437576294, 2.0946145057678223, 0.9222062826156616, 0.7560604810714722, 1.297479510307312, 0.8314876556396484, 0.879172682762146, 1.099845647...
245
[ 0.2885189652442932, 0.4282143712043762, 0.3853171467781067, 0.3179088532924652, 0.3787664473056793, 0.42749640345573425, 0.749443769454956, 0.6373445987701416, 0.9055281281471252, 0.3620363473892212, 0.36998581886291504, 0.6572452783584595, 0.743076741695404, 0.5537398457527161, 0.623874...
247
[ 6.731202602386475, 7.080647945404053, 6.473275184631348, 7.044394016265869, 6.685587406158447, 6.639525890350342, 7.0868120193481445, 6.561972618103027, 6.939984321594238, 6.923398494720459, 6.938863754272461, 6.663183212280273, 7.0552978515625, 6.668303489685059, 6.186368942260742, 6....
282
[ 4.214120388031006, 3.9125053882598877, 3.953235387802124, 3.1412384510040283, 4.9430694580078125, 4.179073333740234, 3.4195995330810547, 3.6455211639404297, 3.9903006553649902, 4.827810287475586, 4.706244945526123, 5.143409252166748, 3.934993028640747, 5.000241279602051, 3.20759415626525...
297
[ 1.8630801439285278, 2.854161262512207, 3.053612470626831, 3.42592716217041, 2.3479104042053223, 2.8890247344970703, 3.2603116035461426, 3.59318208694458, 3.8501434326171875, 2.5288493633270264, 3.0922274589538574, 3.697232246398926, 3.686612129211426, 3.6817030906677246, 3.90573811531066...
327
[ 1.8606401681900024, 2.3841397762298584, 2.271975040435791, 1.7634437084197998, 1.4536566734313965, 1.9231359958648682, 1.7622226476669312, 1.5732388496398926, 1.7041276693344116, 1.823904037475586, 1.431689739227295, 1.9810800552368164, 1.8252739906311035, 1.5040136575698853, 1.750731825...
337
[ 0.00554191367700696, 0.011346101760864258, 0.004937004763633013, 0.06580712646245956, 0.024063315242528915, 0.012018292210996151, 0.014599517919123173, 0.00454559363424778, 0.01248131599277258, 0.013251331634819508, 0.01414499618113041, 0.023418506607413292, 0.015009092167019844, 0.0089144...
376
[ 0.1121918335556984, 0.2619650661945343, 0.20205511152744293, 0.21701884269714355, 0.3471336364746094, 0.3299712538719177, 0.09465064853429794, 0.1733638495206833, 0.1304127722978592, 0.22598813474178314, 0.1101127490401268, 0.19470809400081635, 0.26615211367607117, 0.24706436693668365, 0...
411
[ 0.06775183230638504, 0.04692508280277252, 0.06438913941383362, 0.018792325630784035, 0.035127904266119, 0.024687472730875015, 0.02083790861070156, 0.033724263310432434, 0.02064559981226921, 0.01518393587321043, 0.024353861808776855, 0.03882124647498131, 0.025063058361411095, 0.029039211571...
520
[ 3.828616142272949, 3.675117254257202, 3.1922125816345215, 3.259666919708252, 3.401937961578369, 2.7073724269866943, 2.9589309692382812, 2.2006571292877197, 3.5670862197875977, 3.249037742614746, 2.284823179244995, 3.218698501586914, 2.922232151031494, 2.3534140586853027, 2.72060728073120...
585
[ 4.924289703369141, 5.398392677307129, 3.4795827865600586, 4.895986080169678, 3.9536633491516113, 4.584959983825684, 4.613125801086426, 3.0901944637298584, 4.379059791564941, 2.902865409851074, 3.9049763679504395, 4.260976791381836, 3.385848045349121, 3.2216272354125977, 3.763505935668945...
590
[ 1.0472157001495361, 1.5456773042678833, 2.6290555000305176, 0.6388116478919983, 0.735796332359314, 1.2158294916152954, 0.5789913535118103, 0.5937260985374451, 1.2743444442749023, 1.285975456237793, 0.38659507036209106, 0.6222057342529297, 0.7999405264854431, 0.7779679298400879, 1.4837853...
682
[ 0.1419273167848587, 0.15462341904640198, 0.18716643750667572, 0.1726330667734146, 0.20147502422332764, 0.32150542736053467, 0.2749572694301605, 0.11672050505876541, 0.16116157174110413, 0.22003164887428284, 0.49951159954071045, 0.09413714706897736, 0.2647392749786377, 0.11045342683792114, ...
715
[ 1.2795190811157227, 1.373154878616333, 1.5833224058151245, 1.3146698474884033, 1.013564109802246, 1.6611356735229492, 1.6832499504089355, 1.6322718858718872, 1.7072454690933228, 1.536604642868042, 1.837436318397522, 2.0310473442077637, 1.7656354904174805, 1.9524521827697754, 2.0857112407...
960
[ 0.8474169969558716, 1.0257221460342407, 0.7131783962249756, 0.7796642184257507, 0.8461622595787048, 0.784631609916687, 0.9707021117210388, 0.76091068983078, 0.9786471128463745, 0.6971392631530762, 0.7393771409988403, 0.6932440996170044, 0.6438290476799011, 0.7490140199661255, 0.752649188...
1,148
[ 0.342133492231369, 0.22768254578113556, 0.22826746106147766, 0.2868501842021942, 0.6280534267425537, 0.6841561794281006, 0.5125075578689575, 0.45149508118629456, 0.19779366254806519, 0.3610245883464813, 0.5359125137329102, 0.38959547877311707, 0.324962317943573, 0.277895987033844, 0.1695...
1,321
[ 1.062946081161499, 1.2714918851852417, 0.9482324719429016, 0.809634268283844, 1.1015431880950928, 1.3619304895401, 1.1626321077346802, 0.9551017880439758, 0.6079156398773193, 1.143289566040039, 0.9139626026153564, 1.0577900409698486, 0.6651310324668884, 1.021942377090454, 1.0708707571029...
1,401
[ 0.004114021547138691, 0.007161897141486406, 0.0033053308725357056, 0.010894107632339, 0.004047655034810305, 0.013624651357531548, 0.018083348870277405, 0.008556726388633251, 0.010136672295629978, 0.011285046115517616, 0.004101318307220936, 0.0007917130133137107, 0.008419257588684559, 0.011...
1,404
[ 0.09119629114866257, 0.16752296686172485, 0.127301886677742, 0.1014975979924202, 0.2072475403547287, 0.6676174402236938, 0.328633189201355, 0.22645379602909088, 0.0208111722022295, 0.02975909598171711, 0.1235397681593895, 0.09787456691265106, 0.025599448010325432, 0.03447522222995758, 0....
1,484
[ 0.3883771300315857, 0.4833909273147583, 0.43400102853775024, 0.39820343255996704, 0.6611891984939575, 0.1362767070531845, 0.4613833725452423, 0.3190383315086365, 0.5580059289932251, 0.45861437916755676, 0.8875916004180908, 0.7560579776763916, 0.40421056747436523, 0.49289754033088684, 0.8...
1,510
[ 0.10000273585319519, 0.054025232791900635, 0.01968643069267273, 0.0177219957113266, 0.029221806675195694, 0.07385483384132385, 0.01444877777248621, 0.009183901362121105, 0.014584362506866455, 0.01959291845560074, 0.01613922044634819, 0.010279450565576553, 0.023109516128897667, 0.0047557340...
1,642
[ 0.0043256767094135284, 0.012934358790516853, 0.004250777419656515, 0.006321553606539965, 0.004745886195451021, 0.002862881403416395, 0.004646575544029474, 0.009181184694170952, 0.004002418369054794, 0.0031603651586920023, 0.010323695838451385, 0.0019075546879321337, 0.0037087483797222376, ...
1,902
[ 0.04429606348276138, 0.018970251083374023, 0.25288447737693787, 0.10897383093833923, 0.07208713889122009, 0.051084138453006744, 0.02997904270887375, 0.05785288289189339, 0.09256603568792343, 0.04666093364357948, 0.03844425454735756, 0.06741512566804886, 0.1868666410446167, 0.04468453675508...
1,913
[ 0.02988637238740921, 0.14067693054676056, 0.06615519523620605, 0.072971872985363, 0.0800076425075531, 0.05566231161355972, 0.14280231297016144, 0.016895290464162827, 0.024764694273471832, 0.040401641279459, 0.04501941055059433, 0.019055869430303574, 0.05056510120630264, 0.06677549332380295...
1,943
[ 0.12333166599273682, 0.048128798604011536, 0.03637831285595894, 0.24567368626594543, 0.16993947327136993, 0.07751128077507019, 0.10029533505439758, 0.027714191004633904, 0.24485430121421814, 0.07791374623775482, 0.023359814658761024, 0.04882746934890747, 0.40039485692977905, 0.030843442305...
2,313
[ 0.15163151919841766, 0.16816571354866028, 0.11146718263626099, 0.397777795791626, 0.18088200688362122, 0.3320312798023224, 0.26053810119628906, 0.23938871920108795, 0.36964502930641174, 0.16783426702022552, 0.21342439949512482, 0.128616064786911, 0.14190372824668884, 0.19697421789169312, ...
2,326
[ 1.377568006515503, 1.3311887979507446, 1.5038830041885376, 1.7492700815200806, 1.1315736770629883, 1.720590591430664, 1.3540433645248413, 1.568885326385498, 1.2150663137435913, 1.696236252784729, 1.3834227323532104, 1.7897212505340576, 1.4781827926635742, 1.4143048524856567, 1.6449325084...
2,401
[ 0.01438509114086628, 0.013869103975594044, 0.04682577773928642, 0.02555982396006584, 0.033431123942136765, 0.04802529886364937, 0.020809419453144073, 0.01780877448618412, 0.02835160121321678, 0.018337365239858627, 0.014435265213251114, 0.007940748706459999, 0.02608468197286129, 0.015861295...
2,492
[ 0.17880789935588837, 0.11174258589744568, 0.2418346107006073, 0.28333067893981934, 0.4202944040298462, 0.17075875401496887, 0.18998414278030396, 0.33657532930374146, 0.4725536108016968, 0.17129425704479218, 0.3324255645275116, 0.3566772937774658, 0.26047536730766296, 0.11288262158632278, ...
2,573
[ 0.024082865566015244, 0.020428845658898354, 0.015250388532876968, 0.01590435579419136, 0.00845212023705244, 0.013590429909527302, 0.008301159366965294, 0.0248300489038229, 0.0038073209580034018, 0.010380326770246029, 0.010006265714764595, 0.01113203726708889, 0.028560390695929527, 0.040187...
2,638
[ 1.5185871124267578, 0.3468489944934845, 1.468428611755371, 0.5930702686309814, 0.2238486409187317, 0.31405508518218994, 0.22800490260124207, 0.6316862106323242, 1.4149973392486572, 0.22682519257068634, 0.7049511075019836, 0.2585163712501526, 0.44468504190444946, 0.3136906027793884, 0.381...
2,643
[ 0.011752298101782799, 0.02969973161816597, 0.060084134340286255, 0.021975815296173096, 0.08248747140169144, 0.12116074562072754, 0.05812646821141243, 0.022807423025369644, 0.04860992729663849, 0.017191674560308456, 0.013548916205763817, 0.05162600055336952, 0.11069506406784058, 0.107146926...
2,711
[ 0.011658158153295517, 0.009374291636049747, 0.03110174834728241, 0.01037454605102539, 0.01630307547748089, 0.0028455264400690794, 0.010772173292934895, 0.007753751706331968, 0.01116940751671791, 0.006430292036384344, 0.00957550946623087, 0.003577501280233264, 0.01086934469640255, 0.0049631...
2,860
[ 4.302402496337891, 4.964092254638672, 4.28352165222168, 3.7437219619750977, 3.9531893730163574, 3.562415599822998, 4.716941833496094, 3.366466760635376, 3.9963736534118652, 4.796795845031738, 4.667480945587158, 5.304308891296387, 4.470860481262207, 4.425470352172852, 4.990340232849121, ...
2,955
[ 1.0927863121032715, 0.610457718372345, 0.6462560892105103, 1.480838656425476, 1.5284414291381836, 1.1876493692398071, 1.1187962293624878, 1.0174514055252075, 1.6587661504745483, 1.3232269287109375, 1.5833193063735962, 1.5359210968017578, 1.8075525760650635, 0.7306604385375977, 0.93824690...
3,147
[ 0.04502055048942566, 0.05758102610707283, 0.09326019138097763, 0.12999771535396576, 0.014948612079024315, 0.0908273234963417, 0.07947158813476562, 0.031420525163412094, 0.0027695654425770044, 0.03964664414525032, 0.09272707253694534, 0.016052531078457832, 0.010891631245613098, 0.0512243583...
3,161
[ 0.041017286479473114, 0.022440314292907715, 0.037498340010643005, 0.022939782589673996, 0.028734734281897545, 0.025633957237005234, 0.06247292831540108, 0.06101667135953903, 0.042716674506664276, 0.058563798666000366, 0.054459281265735626, 0.0719420462846756, 0.0407068207859993, 0.02749687...
3,456
[ 0.06364723294973373, 0.041255708783864975, 0.0030644622165709734, 0.00383653468452394, 0.0036618339363485575, 0.006285778246819973, 0.034260522574186325, 0.0049408008344471455, 0.02340325154364109, 0.011757835745811462, 0.29067274928092957, 0.0026120375841856003, 0.01926076039671898, 0.003...
3,502
[ 3.494148015975952, 2.044501304626465, 3.9073328971862793, 2.4516963958740234, 1.3637878894805908, 1.8058950901031494, 1.9283428192138672, 2.9872093200683594, 1.8647034168243408, 2.1266918182373047, 1.1988327503204346, 2.4612505435943604, 2.2088754177093506, 1.2824127674102783, 2.30126881...
3,694
[ 6.149125099182129, 5.096424102783203, 4.700126647949219, 4.954379558563232, 4.547780990600586, 5.1641669273376465, 5.145400047302246, 6.032493591308594, 6.83554744720459, 5.950579643249512, 6.84196662902832, 6.21893310546875, 5.370270729064941, 6.1863861083984375, 5.9240875244140625, 5...
3,707
[ 2.7027506828308105, 3.7852752208709717, 3.452393054962158, 2.3340251445770264, 3.1832573413848877, 2.9987168312072754, 3.26399302482605, 3.974006414413452, 3.834031105041504, 3.4906716346740723, 3.3347203731536865, 4.26708984375, 2.673384189605713, 4.165365219116211, 3.2218596935272217, ...
3,727
[ 2.3290538787841797, 1.6880202293395996, 2.3076300621032715, 2.339261531829834, 1.4858063459396362, 1.5692217350006104, 2.0361430644989014, 2.0429723262786865, 2.925311803817749, 2.1676480770111084, 2.1694986820220947, 1.9859473705291748, 2.0684711933135986, 2.5887389183044434, 1.82431530...
3,745
[ 5.001315593719482, 4.1198296546936035, 2.6369776725769043, 3.6098179817199707, 3.7044034004211426, 4.431476593017578, 4.1739182472229, 3.5820422172546387, 2.2323391437530518, 3.2736616134643555, 3.280425548553467, 3.4152822494506836, 3.4754254817962646, 3.7742867469787598, 3.812791585922...
3,751
[ 2.6107165813446045, 2.166614294052124, 2.5715460777282715, 2.718590259552002, 2.180663585662842, 2.1201162338256836, 2.1951870918273926, 2.2247426509857178, 2.117255449295044, 2.2323901653289795, 1.923448085784912, 2.1637802124023438, 1.7012507915496826, 2.0819735527038574, 2.12820672988...
3,789
[ 0.8423886299133301, 0.8391805291175842, 1.3381669521331787, 0.9707556962966919, 1.0018666982650757, 0.9565978050231934, 1.4016427993774414, 1.0529347658157349, 0.8965342044830322, 1.054162621498108, 0.8980292081832886, 0.5515291690826416, 0.9474608898162842, 0.7440933585166931, 1.0913759...
3,948
[ 0.21757185459136963, 0.4469974637031555, 0.1322343647480011, 0.18315908312797546, 0.22446389496326447, 0.10044337809085846, 0.13206076622009277, 0.0618818961083889, 0.29492121934890747, 0.24944663047790527, 0.3442363440990448, 0.18007704615592957, 0.11304499208927155, 0.10576238483190536, ...
4,093
[ 5.411005020141602, 5.647285461425781, 4.553479194641113, 5.512170791625977, 5.774736404418945, 4.459810256958008, 6.2103047370910645, 4.74212646484375, 5.106485366821289, 5.265491962432861, 5.390171527862549, 5.429478168487549, 5.705745220184326, 4.925114631652832, 4.942688465118408, 4...
4,181
[ 0.07496553659439087, 0.08469372987747192, 0.23454104363918304, 0.2893209159374237, 0.07967063039541245, 0.11539781093597412, 0.09211082011461258, 0.1629379838705063, 0.10221012681722641, 0.06800957769155502, 0.1555636078119278, 0.06712166219949722, 0.18752147257328033, 0.34916234016418457,...
4,240
[ 1.2824549674987793, 2.094053268432617, 2.4085774421691895, 1.407825231552124, 2.1531004905700684, 2.067017078399658, 1.4866759777069092, 1.8624557256698608, 1.5498063564300537, 1.676859974861145, 1.2114595174789429, 1.4377779960632324, 1.4356011152267456, 1.2515517473220825, 1.3588904142...
4,376
[ 1.6796108484268188, 2.150247573852539, 2.8038392066955566, 2.2277204990386963, 1.9541938304901123, 2.147237777709961, 2.1513805389404297, 1.6833255290985107, 2.157256603240967, 2.648350715637207, 2.8582329750061035, 1.7656910419464111, 2.840790271759033, 3.164581775665283, 2.512567043304...
4,444
[ 3.2898166179656982, 3.221604824066162, 2.216012477874756, 3.126105308532715, 2.515582799911499, 2.496328353881836, 3.1543993949890137, 4.098471641540527, 3.904231309890747, 3.5667688846588135, 3.3141674995422363, 4.03322172164917, 2.573120594024658, 3.371591567993164, 3.1585538387298584,...
4,685
[ 0.9511681795120239, 1.3327960968017578, 1.3786792755126953, 1.205521583557129, 0.8144357204437256, 1.1417889595031738, 0.7999561429023743, 1.1233525276184082, 0.8095137476921082, 1.0407438278198242, 0.8676265478134155, 1.1542704105377197, 1.1491833925247192, 1.1357613801956177, 1.2457450...
4,727
[ 0.6567894220352173, 0.591675877571106, 0.8246360421180725, 1.0343554019927979, 1.1666992902755737, 1.2714824676513672, 1.3004882335662842, 1.487173318862915, 0.8955109119415283, 1.1096293926239014, 1.5158182382583618, 0.9860430359840393, 1.1110856533050537, 1.1910147666931152, 1.66316175...
4,854
[ 4.557668209075928, 4.400118827819824, 4.282637119293213, 4.873676300048828, 4.467620372772217, 4.6901445388793945, 4.610644340515137, 4.840844631195068, 4.735603332519531, 4.796436309814453, 4.533334255218506, 4.520535945892334, 4.715275287628174, 4.58407735824585, 4.677913665771484, 4...
4,894
[ 0.011319464072585106, 0.0037486536893993616, 0.01782095432281494, 0.01322639174759388, 0.009732888080179691, 0.011364017613232136, 0.012111691758036613, 0.023019935935735703, 0.008531669154763222, 0.01653221808373928, 0.0059007019735872746, 0.007997987791895866, 0.012895407155156136, 0.011...
4,961
[ 0.487056702375412, 0.5601791143417358, 0.4470441937446594, 0.5803607702255249, 0.5569785237312317, 0.41185304522514343, 0.41909125447273254, 0.45880892872810364, 0.3950023949146271, 0.40812644362449646, 0.47018736600875854, 0.48368361592292786, 0.42974090576171875, 0.34586331248283386, 0...
5,220
[ 4.03061580657959, 3.663184642791748, 3.415174961090088, 2.976750612258911, 3.3778281211853027, 3.319136142730713, 3.4286038875579834, 3.401125431060791, 3.5773589611053467, 3.3999743461608887, 3.4070639610290527, 2.8597874641418457, 3.4969310760498047, 3.8965039253234863, 3.3408980369567...
5,258
[ 0.052636463195085526, 0.02272200956940651, 0.05219574272632599, 0.008269711397588253, 0.01844492182135582, 0.02730248123407364, 0.06805123388767242, 0.09156892448663712, 0.1302449256181717, 0.02108772285282612, 0.026525335386395454, 0.027080317959189415, 0.07314340770244598, 0.030169680714...
5,299
[ 6.783290863037109, 6.740674018859863, 6.343559265136719, 6.354007244110107, 7.12945556640625, 7.2280120849609375, 6.544414520263672, 6.959272384643555, 6.808381080627441, 6.909539699554443, 6.731292247772217, 7.355398654937744, 6.68342399597168, 6.671435832977295, 6.439164638519287, 6....
5,307
[ 5.472689151763916, 5.825456619262695, 5.632530689239502, 4.814324378967285, 5.172332763671875, 5.674750328063965, 5.246674537658691, 5.373987197875977, 4.945071220397949, 5.356806755065918, 5.1346588134765625, 4.605759620666504, 5.937788009643555, 4.447115421295166, 4.483432292938232, ...
5,658
[ 5.98388671875, 5.7742815017700195, 6.235109806060791, 5.570125579833984, 6.34799861907959, 6.938157081604004, 6.040212631225586, 6.083623886108398, 6.014888763427734, 6.176970481872559, 6.421292781829834, 6.1115193367004395, 6.444275379180908, 6.13095235824585, 6.617161750793457, 5.874...
5,786
[ 0.5180702209472656, 0.621735692024231, 0.974714994430542, 0.9220014810562134, 0.4743727147579193, 0.8048076629638672, 1.623949646949768, 0.9212157726287842, 1.224243402481079, 0.6995188593864441, 0.6460040807723999, 1.1127986907958984, 0.8989871144294739, 0.5469052791595459, 0.6956205964...
5,829
[ 9.386344909667969, 9.207636833190918, 10.075437545776367, 8.51812744140625, 9.601563453674316, 9.08850383758545, 8.769604682922363, 9.337583541870117, 8.568288803100586, 9.809147834777832, 8.390230178833008, 9.10531997680664, 8.57650375366211, 9.076626777648926, 8.626640319824219, 8.02...
5,949
[ 4.976224899291992, 3.8995940685272217, 4.950814247131348, 5.753294467926025, 4.589847564697266, 4.217694282531738, 4.116024017333984, 5.308101654052734, 5.0412516593933105, 4.620273590087891, 5.662456035614014, 4.572612762451172, 4.6098761558532715, 5.1991143226623535, 3.988582134246826,...
6,122
[ 7.900481224060059, 7.103255271911621, 7.910953998565674, 7.6541290283203125, 8.286243438720703, 7.474429607391357, 8.005010604858398, 7.681417942047119, 6.225075721740723, 7.946310043334961, 7.453945636749268, 7.346959114074707, 7.257328987121582, 6.574943542480469, 7.489841461181641, ...
6,280
[ 0.01075141690671444, 0.021147483959794044, 0.05170094221830368, 0.03666817769408226, 0.02053956501185894, 0.08172350376844406, 0.059862736612558365, 0.05129151791334152, 0.017996594309806824, 0.017573483288288116, 0.016161741688847542, 0.02661263197660446, 0.013788331300020218, 0.015706637...
6,314
[ 1.6710366010665894, 1.5334118604660034, 2.0783233642578125, 2.182494878768921, 1.6855374574661255, 2.1073198318481445, 1.5252676010131836, 2.1259818077087402, 2.03399658203125, 1.9628199338912964, 1.991461157798767, 1.5944383144378662, 1.619309902191162, 1.7142714262008667, 1.90062069892...
6,347
[ 4.537757873535156, 4.526512145996094, 4.723887920379639, 3.9285101890563965, 4.906861305236816, 4.315976142883301, 5.059999465942383, 5.004002571105957, 4.6365532875061035, 4.609781265258789, 4.741829872131348, 4.7702741622924805, 4.559216022491455, 4.699963569641113, 4.961742401123047, ...
6,415
[ 0.43517324328422546, 0.6381569504737854, 0.7039697170257568, 0.5767594575881958, 0.6959742903709412, 0.4510156512260437, 0.5042695999145508, 0.519322395324707, 0.817195475101471, 0.4846767485141754, 0.5773822665214539, 0.4871780574321747, 0.5002927780151367, 0.3139292299747467, 0.4580762...
6,540
[ 7.591266632080078, 7.008116722106934, 5.9070940017700195, 6.640912055969238, 6.9431610107421875, 7.8423638343811035, 6.963910102844238, 7.456160068511963, 7.18684720993042, 6.257257461547852, 6.888352870941162, 7.368887424468994, 7.358631134033203, 8.18078327178955, 7.366121292114258, ...
6,639
[ 4.3720316886901855, 4.40142822265625, 4.581624507904053, 4.219712257385254, 5.129771709442139, 4.6958465576171875, 3.548523426055908, 4.575671195983887, 4.727945327758789, 4.780162811279297, 5.0540924072265625, 4.873059272766113, 5.480872631072998, 4.4650373458862305, 4.944950103759766, ...
6,748
[ 1.024446725845337, 1.266839861869812, 1.3024673461914062, 0.9945589900016785, 1.2096595764160156, 1.8953112363815308, 1.0254650115966797, 1.1761903762817383, 1.2280285358428955, 1.2042815685272217, 0.8321214914321899, 0.8402091860771179, 0.9293339252471924, 0.7495740056037903, 0.57687222...
6,835
[ 1.2748397588729858, 1.2185866832733154, 1.544159173965454, 1.6062029600143433, 1.7228022813796997, 1.6617977619171143, 1.6712175607681274, 1.4317293167114258, 1.691603422164917, 1.6502727270126343, 1.7197670936584473, 1.4874708652496338, 1.5481308698654175, 1.6476362943649292, 1.60334610...
6,979
[ 0.696458101272583, 0.809978187084198, 0.8685930371284485, 0.8513251543045044, 0.6739380359649658, 0.7603747248649597, 0.6589750051498413, 0.8224382996559143, 0.7991368770599365, 0.638853132724762, 0.7756605744361877, 0.8052477836608887, 0.7668484449386597, 0.8059722781181335, 0.772957146...
7,043
[ 3.3415441513061523, 3.1304450035095215, 3.2368974685668945, 3.044466018676758, 2.8509364128112793, 2.994600296020508, 3.1080875396728516, 3.1183786392211914, 3.271615982055664, 2.9267044067382812, 2.991533041000366, 3.1597237586975098, 3.0161733627319336, 3.1910810470581055, 2.9594364166...
7,155
[ 10.289902687072754, 9.18813419342041, 10.86448860168457, 9.64404296875, 10.624931335449219, 9.451558113098145, 10.10946273803711, 11.676460266113281, 10.807670593261719, 10.81612777709961, 8.775589942932129, 10.602251052856445, 11.032670021057129, 11.561028480529785, 11.964008331298828, ...
7,179
[ 1.8008081912994385, 1.981600284576416, 2.3394203186035156, 2.4799790382385254, 2.272686004638672, 1.7662699222564697, 1.8105905055999756, 1.7381874322891235, 2.282104015350342, 2.2289490699768066, 1.8008369207382202, 2.0696001052856445, 2.069549083709717, 2.011317729949951, 2.15328240394...
7,191
[ 7.276383399963379, 6.389397144317627, 6.686855792999268, 7.518772125244141, 7.058168411254883, 6.855903148651123, 6.956897735595703, 6.6532182693481445, 6.978373050689697, 6.034763336181641, 6.770238876342773, 7.136950492858887, 7.765472412109375, 6.844181060791016, 6.922668933868408, ...
7,448
[ 8.040863037109375, 7.3427581787109375, 7.689668655395508, 8.628643035888672, 8.211884498596191, 8.952588081359863, 8.440622329711914, 7.680253028869629, 7.391432762145996, 7.847197532653809, 8.574919700622559, 7.482348918914795, 7.830106735229492, 8.171791076660156, 8.540290832519531, ...
7,679
[ 7.1761369705200195, 6.770722389221191, 6.8020477294921875, 7.460542678833008, 7.139777183532715, 6.926537036895752, 7.400941848754883, 7.195428371429443, 6.705295562744141, 7.310667037963867, 7.047211170196533, 7.836945533752441, 7.3597283363342285, 6.911499977111816, 7.992453575134277, ...
7,954
[ 0.6477296352386475, 0.5763073563575745, 0.6810747385025024, 0.5446444153785706, 0.6002101898193359, 0.489975243806839, 0.5428211688995361, 0.5757195949554443, 0.5622137784957886, 0.619459867477417, 0.7152057886123657, 0.6971558928489685, 0.8307342529296875, 0.6727267503738403, 0.74645185...
7,958
[ 3.067007541656494, 4.026584625244141, 4.104979038238525, 4.11241340637207, 3.0501554012298584, 3.370851755142212, 3.4966089725494385, 3.2973971366882324, 3.1357455253601074, 2.512882709503174, 3.550853729248047, 2.720351219177246, 3.855194091796875, 4.095219612121582, 3.644775390625, 3...
8,012
[ 10.522895812988281, 10.725192070007324, 10.65351390838623, 11.275923728942871, 10.388742446899414, 11.069221496582031, 11.94274616241455, 11.712859153747559, 11.234787940979004, 12.53547477722168, 11.642690658569336, 11.335803985595703, 11.946654319763184, 11.029425621032715, 11.34494400...
8,015
[ 1.198617696762085, 1.4263978004455566, 1.0855871438980103, 1.0801221132278442, 1.2177361249923706, 1.3770726919174194, 1.2700119018554688, 1.0673549175262451, 1.1900585889816284, 1.2913845777511597, 1.0607649087905884, 1.2050849199295044, 1.145703673362732, 1.1810656785964966, 1.40011250...
8,045
[ 2.926046133041382, 2.759495496749878, 3.3010363578796387, 2.616955280303955, 3.4062607288360596, 3.1051249504089355, 2.5486319065093994, 2.87127947807312, 2.67071270942688, 3.1522316932678223, 2.8012428283691406, 2.4198527336120605, 3.0776138305664062, 3.0722076892852783, 2.9177968502044...
8,153
[ 0.4820851683616638, 0.7916881442070007, 0.3841027319431305, 0.36583346128463745, 0.37045902013778687, 0.5911438465118408, 0.4568830132484436, 0.28866249322891235, 0.21203896403312683, 0.562640905380249, 0.2714245915412903, 0.2962852716445923, 0.8395281434059143, 0.5110317468643188, 0.345...
8,201
[ 3.2550296783447266, 3.045968532562256, 2.8802196979522705, 3.1560468673706055, 3.355724334716797, 2.3751935958862305, 3.1037497520446777, 3.1368508338928223, 3.1752920150756836, 2.9202823638916016, 3.3911945819854736, 3.4391419887542725, 3.1734795570373535, 3.2286770343780518, 3.54712891...
8,286
[ 6.318672180175781, 6.7985076904296875, 8.209725379943848, 8.052003860473633, 7.502462863922119, 8.913901329040527, 8.379375457763672, 7.64036750793457, 8.039655685424805, 8.2244873046875, 8.451136589050293, 8.857156753540039, 8.806148529052734, 7.332640171051025, 8.71833610534668, 8.42...
8,335
[ 7.960304260253906, 7.359225273132324, 7.406723976135254, 7.185488224029541, 8.229759216308594, 7.456808090209961, 7.585362434387207, 8.152213096618652, 8.33371639251709, 8.093670845031738, 7.069888114929199, 6.823494911193848, 7.31278657913208, 6.724331855773926, 7.8184709548950195, 6....
8,430
[ 0.31262072920799255, 0.5360586643218994, 0.3705696165561676, 0.39034685492515564, 0.2801908552646637, 0.3646448850631714, 0.3496580123901367, 0.4098740518093109, 0.4076147675514221, 0.4445289671421051, 0.3724737763404846, 0.31955650448799133, 0.5265365839004517, 0.39234861731529236, 0.41...
8,578
[ 2.642399549484253, 2.38730525970459, 2.3215978145599365, 2.3014838695526123, 2.42838716506958, 2.3434081077575684, 2.768199920654297, 1.4226380586624146, 1.515638828277588, 0.9461498260498047, 1.9734594821929932, 1.764025092124939, 2.2386739253997803, 1.8312509059906006, 3.11767578125, ...
8,611
[ 0.17493750154972076, 0.23610404133796692, 0.18646614253520966, 0.3563997149467468, 0.23693861067295074, 0.16157911717891693, 0.19180212914943695, 0.19676430523395538, 0.27817267179489136, 0.3021802306175232, 0.247394859790802, 0.2959878444671631, 0.25773850083351135, 0.20649467408657074, ...
8,682
[ 2.491722345352173, 2.535750150680542, 1.8822880983352661, 2.963041305541992, 1.6429697275161743, 1.8647081851959229, 1.6887483596801758, 1.286712646484375, 1.5964891910552979, 2.4845709800720215, 1.7551214694976807, 1.7996468544006348, 1.9375643730163574, 2.578965663909912, 1.36955857276...
8,790
[ 10.232277870178223, 10.481954574584961, 9.243570327758789, 10.40575885772705, 9.770816802978516, 9.150676727294922, 9.580211639404297, 10.008583068847656, 9.696541786193848, 9.769428253173828, 9.385309219360352, 9.422432899475098, 9.983097076416016, 10.195977210998535, 9.988504409790039,...
8,884
[ 5.715084552764893, 6.747357368469238, 6.129428386688232, 5.6002607345581055, 5.717398166656494, 6.0524492263793945, 7.176349639892578, 6.591335773468018, 6.075819969177246, 5.473771095275879, 6.145131587982178, 4.6550679206848145, 5.980997085571289, 5.670500755310059, 5.291577339172363, ...
8,914
[ 3.1715340614318848, 4.107014179229736, 3.919584035873413, 4.1432647705078125, 3.193188428878784, 3.878230094909668, 2.890042543411255, 3.254469156265259, 2.901516914367676, 3.3282549381256104, 2.749030590057373, 3.935899496078491, 2.510735273361206, 3.0484538078308105, 3.2932229042053223...
9,091
[ 2.013984203338623, 1.9635496139526367, 1.801444172859192, 1.7092952728271484, 1.6935092210769653, 1.871891736984253, 2.2002615928649902, 1.7694284915924072, 1.7681798934936523, 2.4728450775146484, 1.8256205320358276, 2.1587798595428467, 2.1011006832122803, 2.0395922660827637, 1.730815887...
End of preview. Expand in Data Studio

Smooth Scaling Laws Hide Stepwise Token Learning, NeurIPS 2026 Spotlight. Each token's loss falls in one sigmoid-shaped step, and their average is the smooth loss curve.

Interactive project page Paper on OpenReview Code on GitHub Quickstart in Colab BibTeX

Stepwise Token Learning: Token-Level Loss Trajectories

Per-token validation-loss trajectories for one million contextualized tokens along each of three scaling axes of LLM pre-training (training steps T, data scale D, model scale M), plus a synthetic arithmetic example. They are the inputs to the analysis in Smooth Scaling Laws Hide Stepwise Token Learning (NeurIPS 2026, Spotlight).

See these trajectories move. The interactive project page plots tokens from these matrices, shows how they add up to the loss curve, and lets you play with the mechanism. The code on GitHub downloads these files and reproduces every figure with one command.

What the Matrices Show

Each row is one contextualized token, and each column is a checkpoint or a run along the scaling axis a. Individual rows drop in localized, sigmoid-like steps, which the paper fits with one learning event per token:

β„“i(a)β‰ˆβ„“ipost+Ξ”i[1βˆ’Οƒ(Ξ»i(aβˆ’Ο„i))] \ell_i(a) \approx \ell_i^{\mathrm{post}} + \Delta_i \left[ 1 - \sigma\big( \lambda_i (a - \tau_i) \big) \right]

Here Ο„ is when the token is learned, Ξ» how abruptly, and Ξ” how much loss it sheds. The learning pulses share one shape g, so the slope of the averaged loss L is the distribution of learning times, the learning-time spectrum p(Ο„), smoothed by g:

βˆ’dLdaβ‰ˆβˆ«p(Ο„) g(aβˆ’Ο„) dΟ„ -\frac{\mathrm{d}L}{\mathrm{d}a} \approx \int p(\tau) \thinspace g(a - \tau) \thinspace \mathrm{d}\tau

Token-level loss trajectories and the learning-time spectrum along training steps, data scale, and model scale.

Contents

Asset Scaling axis Rows Γ— points Axis range NPZ size Paper Viewer config
step_axis_1m Training steps T 1,000,000 Γ— 54 15,000–71,000 steps 196.7 MB Figure 3 step_axis_1m (10K-row preview)
d_axis_1m Data scale D 1,000,000 Γ— 50 1.196–34.662B tokens 183.0 MB Figure 4 d_axis_1m (10K-row preview)
m_axis_1m Model scale M 1,000,000 Γ— 60 0.416–4.999 GFLOPs/token 219.3 MB Figure 5 m_axis_1m (10K-row preview)
synthetic_mano_uniform_validation_10k Training steps 10,000 Γ— 80 5,000–400,000 steps 1.5 MB Appendix synthetic_mano_uniform_validation_10k (all rows)

The full matrices are the losses.npz files. The Dataset Viewer shows Parquet previews: a seeded random sample of 10,000 rows from each T/D/M matrix, the complete synthetic matrix, and an axes table with the axis values of every asset. Byte sizes and SHA256 checksums are listed in manifest.yaml.

Quick Start

Browse a preview with πŸ€— Datasets:

from datasets import load_dataset

repo = "applewpj/stepwise-token-learning"
preview = load_dataset(repo, "step_axis_1m", split="preview")   # columns: row, loss[54]
axes = load_dataset(repo, "axes", split="full")                 # axis values per asset

Load a full matrix with NumPy:

import numpy as np
from huggingface_hub import hf_hub_download

path = hf_hub_download(
    repo_id="applewpj/stepwise-token-learning",
    repo_type="dataset",
    filename="examples/step_axis_1m/losses.npz",
)
with np.load(path, allow_pickle=False) as data:
    steps = data["axis_values"]        # (54,) training steps
    losses = data["loss_matrix"]       # (1000000, 54) float32, one row per token

mean_loss = losses.mean(axis=0)        # the aggregate validation-loss curve

Reproduce the paper figures from these files with the analysis code. Each command downloads the matrix it needs, verifies its checksum, fits it, and checks the result against reference numbers:

git clone https://github.com/applewpj/stepwise-token-learning
cd stepwise-token-learning
pip install -e .
tokenlearning reproduce fig3 --quick   # Figure 3 from 50,000 tokens, about a minute on CPU
tokenlearning reproduce all            # Figures 3-5 and the appendix at the paper setting

The Colab quickstart runs the same pipeline in the browser.

Data Structure

Each losses.npz contains:

axis_values: float64 [K]     ordered positions along the scaling axis
loss_matrix: float32 [N, K]  loss of token row i at axis point j
axis_name:   string  [1]     e.g. "training step"

Rows are aligned: loss_matrix[i, j] is the same contextualized token at every axis point j, which is what makes token-level trajectories meaningful. The synthetic matrix adds per-row metadata: sample_id, token_pos, difficulty, seed, and question (the arithmetic expression).

The Parquet previews have one row per token with columns row (index into the full loss_matrix), any per-row metadata, and loss (the trajectory as a list of K floats). The axes config has one row per asset with asset, axis_name, n_points, n_tokens, preview_rows, and axis_values.

Provenance

Asset How it was produced
step_axis_1m A 1.5B dense model trained on 300B tokens of a mixed corpus (Chinese, English, mathematics, reasoning, books, papers, code); 54 checkpoints.
d_axis_1m A 290M MoE model; a 50-run sweep over training budgets from 1B to 35B tokens, each run with a complete warmup-constant-decay schedule, used to construct the data-scale frontier.
m_axis_1m A 60-run sweep of model variants from 0.42 to 5.00 GFLOPs/token (β‰ˆ400M–6B parameters), each trained on 10B tokens and tuned toward a near-optimal model-scale frontier.
synthetic_mano_uniform_validation_10k The public synthetic arithmetic (Mano) pipeline adapted from PhysicsLM4; 10,000 validation samples scored at 80 checkpoints.

The T/D/M losses are measured on held-out validation tokens; for the step axis, the validation set is sampled from the same distribution as the training corpus. The matrices are sanitized analysis artifacts: they keep aligned token losses and axis values only, and do not include model architectures, weights, tokenizers, validation text, training-data composition, experiment registries, or checkpoint paths.

Uses and Limitations

Suited for reproducing the paper's analysis, testing alternative learning-time estimators or curve-fitting methods, and studying token-level learning dynamics along scaling axes.

  • Rows of the T/D/M matrices cannot be mapped back to token strings or text.
  • These are loss measurements, not a language-model training corpus.
  • Per-token trajectories are noisy; the paper's sigmoid fits describe the dominant transition, not every fluctuation.

Citation

@inproceedings{wang2026smooth,
  title     = {Smooth Scaling Laws Hide Stepwise Token Learning},
  author    = {Wang, Pingjie and Hu, Zechen and Yang, Peiru and Guo, Fu and Zhang, Debing},
  booktitle = {Advances in Neural Information Processing Systems},
  year      = {2026},
  url       = {https://openreview.net/forum?id=THZDVZArRH}
}

License and Contact

The matrices are released under the MIT License. Questions and issues are welcome on the GitHub issue tracker.

Downloads last month
323